id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_wcurz4s52va3rcxqzojkxfwgx4	�andor Dembitz	Developing a Spell Checker as an Expert System	2003	7	.pdf	application/pdf	4614	502	73	concept of energy, not in use in knowledge engineering checker for Croatian language, implemented as learning Keywords: spell checker, knowledge acquisition, modelling of learning, energy. managed to put more than 50% of unknown English word types in the class of the lowest peculiarity; the best result in Croatian was around to bring majority of nonword types at the beginning of Hascheck report; legal Croatian word As mentioned earlier, Hascheck started to operate with a dictionary of less than 100,000 common word types. In the first 5 years its knowledge grew to more than 300,000 Croatian common word types, and more than 80,000 name Hascheck learns words to improve its covering of Croatian texts. as processing of a certain amount of text in order to acquire basic Croatian word knowledge. The result of fitting is presented in Fig. 1, together with optimal values of free function parameters a, ∆t and τ , and with obtained correlation coefficient r.	./cache/work_wcurz4s52va3rcxqzojkxfwgx4.pdf	./txt/work_wcurz4s52va3rcxqzojkxfwgx4.txt